The problem
Vectra came from thinking about what I could build next and looking at the wearables I already use. Most wearables collect information or show notifications. I liked the idea that one could also become a controller, using gestures to interact with the room in the way Tony Stark controls technology around him.
It is not a response to an urgent problem so much as a fun question: could pointing at a lamp, speaker, or screen feel like a natural way to choose it?
What I built
The idea is still theoretical. One possible first prototype would use fixed UWB anchors to locate a wrist tag and an orientation sensor to estimate where the hand is pointing. A local hub could compare that direction with known devices in the room and identify the most likely target.
I would begin with a deliberate confirmation button rather than automatic gestures. That would separate two questions: whether the system can identify the intended object, and whether it can recognise a gesture reliably. If basic pointing worked, gesture recognition and more expressive controls could come later.
Decisions and tradeoffs
Prove point-and-press before gestures
A first prototype would use a physical button for confirmation. Gesture recognition could come later, after the basic spatial selection is shown to work.
One wrist tag, measured orientation
The proposed setup uses four room anchors to locate one wrist tag, with an IMU providing direction. It is only an initial architecture and would need to be tested against simpler alternatives.
Prove the interaction before polishing the object
Custom electronics and an enclosure would only make sense after a rough prototype showed that pointing is accurate and that accidental activations can be controlled.
